Light olefins are an important basic chemical feedstock for the production of many plastics used in a variety of industries. Olefins are most commonly produced from petroleum feedstocks through the cracking of larger hydrocarbon molecules. The cracking process is either a catalytic or steam cracking process, and produces light olefins which consist primarily of ethylene and propylene.
An alternate source of light olefins is from the conversion of oxygenates to olefins. The primary oxygenate that is converted to an olefin is methanol. The preferred process is generally referred to as methanol-to-olefins (MTO) process. The primary olefins produced from this process are ethylene and propylene, and the process is performed over a catalytic molecular sieve. The MTO process enables an important alternative to petroleum sources of feeds for the production of light olefins. The sources of oxygenates include alcohols, such as methanol and ethanol; ethers, such as dimethyl ether and diethyl ether; and other oxygenates, such as methyl formate and dimethyl carbonate. These oxygenates can be produced from natural gas, fermentation of biomass, municipal wastes, and recycled organic materials. An important commercial consideration is that methanol can be readily produced from natural gas, or coal, and is easier and safer to handle and transport then either natural gas or coal.
There are many studies of molecular sieves for the use in methanol to olefin processes, with SAPO-34 disclosed as a preferred molecular sieve. In trying to improve the characteristics of SAPO-34, the molecular sieve has been subjected to various treatments. For example U.S. Pat. No. 5,932,512 discloses that the molecular sieve is synthesized and then treated with a fluoride compound to form a fluorinated silico-aluminophosphate molecular sieve. While there is some improvement in the total selectivity of ethylene plus propylene, there is also a shift in favor of greater ethylene selectivity and lower propylene selectivity. It is important to note that the '512 patent deals with a method of treating an already formed molecular sieve and loading it with fluoride, rather than producing the desired molecular sieve from a synthesis reaction mixture.
In searching for improved SAPO-34 catalysts, Y. Xu, et al., J. Chem. Soc. Faraday Trans. 86(2), 425-429 (1990) studied the effect of hydrofluoric acid (HF) on crystal growth, but found that the presence of HF favors the formation fewer and larger crystals. In addition, Xu et al., used higher concentration of the organic templating agent. The use of a fluoride source is known and discussed in the production of molecular sieves. U.S. Pat. No. 4,786,487 uses a fluoride source, but for the generation of sodalite and SAPO-20 which has an SOD framework type. Different types of molecular sieves are produced under different conditions, and there is no guidance as to the applicability of this to other molecular sieves. However, in the formation of catalyst for use in oxygenates to olefins it is preferred to form smaller crystals as larger crystals reduce the efficiency and shorten the regeneration cycle of the catalyst.
The preparation of a silicoaluminophosphate composition with a CHA framework structure in the presence of fluoride is reported in the PhD dissertation of Erling Halvorsen (K.-P. Lillerud, thesis advisor; University of Oslo, Department of Chemistry, 1996). This material is designated UiO-S4. In this work the authors claim that the preparation of pure UiO-S4 requires a TEAOH/Al2O3 ratio of 2 and low to medium HF content (HF/Al2O3=0.15−0.7). At lower TEAOH levels, mixtures of SAPO-5, SAPO-34, UiO-S6, and UiO-S4 were formed. For pure UiO-S4 the gel composition 2.0 TEAOH.1.0 Al2O3.1.0 P2O5.0.1 SiO2.0.2 HF.50H2O was digested at 150° C. for 21 hours. Halvorsen indicated that the XRD pattern of the as-synthesized product did not resemble the pattern of SAPO-34 much, but upon calcination the familiar pattern of SAPO-34 is observed. Elemental analysis of the product showed (mole fraction, normalized) P 0.461, Al 0.499, Si 0.032, F 0.08. The average crystallite size was approximately 1.0 micrometer. While presenting new materials, including the production of SAPO-34 in a mixture of materials, there was no testing for use in the MTO process.
The synthesis of SAPO-34 with TEAOH and HF is shown in U.S. Pat. No. 5,096,684. Example 10 uses gel composition 1.0 TEAOH.0.6 SiO2.1.5 Al2O3.0.7 P2O5.100H2O .1.0 HF and SEM analysis of the product shows nearly cubic crystals 2-15 μm in size, and a composition of Si0.13Al0.49P0.38. Example 11 uses gel composition 1.0 TEAOH.1.00 SiO2.1.75 Al2O3.0.75 P2O5.100H2O.1.0 HF and produces a product with composition Si0.12Al0.50P0.38. These preps are characterized by high SiO2 levels in the reaction media and generate large crystals for the products.
A method of synthesizing aluminophosphate and silicoaluminophosphate molecular sieves and in particular to the synthesis of aluminophosphate and silicoaluminophosphate molecular sieves using N-methylethanolamine (MEA) as template with or without a source of fluoride is described in U.S. Pat. No. 6,767,858 B1. In example 1 the use of N-methyl ethanolamine as sole template results in good quality SAPO of CHA framework type but with 1.94 Si/CHA cage (16 mol % Si). In example 2, by combining TEACl with the MEA, SAPO with 0.96 SI/CHA cage (8 mol % Si) was produced. Alternatively, in example 3, by incorporating a F-source with the MEA, a SAPO molecular sieve of CHA framework type with an even lower level of silicon, 0.31 Si/CHA cage (2.5 mol % Si) was prepared. In all cases, no indication of crystal size or particle size was given. In the later case the equimolar Al and P content suggested that the acid site density would be very low which will produce a poor MTO catalyst, and not be an improvement.
The current state of production of a suitable SAPO-34 is still fraught with problems such as the generation of intergrowths in the crystals that results in crystals with a structure that has part CHA framework type and part AEI framework type, or the production of crystals that are too large. The mixture of framework types in the catalyst lowers the selectivity, and crystals that are too large rapidly coke up and have reduced activity in the process of methanol to olefin conversion.